Gaming the Laws of War


Review of ORDER WITHIN ANARCHY: The Laws of War as An International Institution, by James D. Morrow

Cambridge University Press, 2014


Beginning with the 1907 Hague Convention, international treaties have regulated how countries are required to treat prisoners of war (POWs). These treaties have required governments to provide POWs with a minimum standard of care — including access to food, shelter, clothing, and relief services.

Countries have not always complied with these obligations. Germany had ratified the relevant treaties regulating the treatment of POWs, but there was a wide discrepancy in how Germany treated POWs on the western and eastern fronts during World War II. While only about 100,000 captured soldiers from Western powers died in German POW camps (4% of the population), over 3,000,000 captured Russian soldiers died in German camps (57% of the population) (pp. 206-08).

The fact that Germany could treat POWs on opposite fronts so differently during the same conflict is staggering. And Germany is not the only country with wide discrepancies in treatment of POWs during the same war. The allied powers — who had also ratified the relevant treaties — were willing to take prisoners in Europe but were less willing to do so in the Pacific theatre (p. 224). As these examples illustrate, predicting whether a country is likely to treat POWs humanely is more complicated than just knowing whether that country has ratified the relevant treaties. 

The treatment of POWs is just one of many topics covered by international treaties that regulate the conduct of war. Treaties also regulate the treatment of civilians, the use of chemical and biological weapons, and conduct of war on the high seas. Taken together, these treaties form what is known as either the laws of war or, more formally, International Humanitarian Law. And just like with the treatment of POWs, countries that have ratified treaties on the laws of war do not always consistently live up to their commitments.

Despite the importance of the topic, there has been little empirical research on whether countries comply with the laws of war—until now. James Morrow’s book is the most comprehensive empirical exploration of when the laws of war succeed at constraining state behavior. In the book, Morrow uses game theory to generate predictions of how states are likely to behave during conflicts, and then tests these predictions using an original dataset of compliance with the laws of war.

*          *          *

Morrow begins by arguing that the laws of war are an “international institution.” International institutions, according to Morrow, provide the “rules of the game” and make interactions regular and predictable. Morrow then draws on game theory to conceive of countries as strategic actors trying to determine their optimal strategy given a certain set of motivations and incentives. For actors to make strategic decisions on how to behave, they need to have a shared understanding of how the other actors in the game will behave.

During conflicts, the laws of war provide that shared understanding. The preferences that countries express through the negotiation and ratification of these international treaties give insight into what restrictions on war that they are likely accept. In other words, the laws of war both clarify what is expected of countries and provides insight into whether a country is likely to comply. In this way, the laws of war can alter the way that countries choose to behave during conflicts. 

Using this starting point, Morrow generates several hypotheses about when states are likely to comply with the laws of war. Most notable is the hypothesis that overall compliance and the correlation of compliance for a given issue area — say treatment of POWs — is likely to be highest when both countries in a conflict have ratified the relevant treaties. The reason for this is that “[w]hen both sides ratify the relevant treaty, they create the shared expectation that one another will comply up to the limits of their control” (p. 87). He also argues that issue areas with decentralized control or higher monitoring costs are likely to have lower average rates of compliance; that first violations are likely to occur early in wars; and that when the first violation comes late in the war the side winning the war is more likely to commit it.[1]

Morrow’s dataset consists of all interstate wars from the Boxer Rebellion in 1899 to the Gulf War in 1991.[2] For each conflict, Morrow breaks out pairs of warring states. For example, during World War II, the United States and Japan are one pair of warring states, and the United States and Germany are another. For nine different issue areas governed by the laws of war — like the treatment of prisoners and protection of civilians — Morrow codes the level of compliance with the laws of war for each country in the warring pair.

Through both quantitative and qualitative analysis, Morrow finds support for his hypotheses. Most important, Morrow finds that joint ratification of treaties on the laws of war between two countries increases the likelihood that they will restrain themselves during conflict, while also increasing the likelihood that there will reciprocal violations when restraint fails. Based on these findings, Morrow concludes that although the laws of war do not dictate the ways that countries fight, they do shape the ways that countries behave because of the shared expectations that they create.

This argument is most clearly illustrated by Germany’s disparate treatment of western and Russian POWs. The western powers and Germany had both ratified the relevant treaties on POWs; and since the western countries indicated there willingness to treat prisoners humanely, Germany reciprocated. Since Russia had not ratified the relevant treaties and did not indicate a willingness to comply with their requirements, Germany did not treat Russian POWs consistently with its legal commitments.

*          *          *

Morrow’s biggest contribution is his methodology. He uses game theory to generate simple hypotheses about the laws of war and then he tests those hypotheses using data. This straightforward style of social science is all too rare in the field of international law. Testing the ideas that Morrow generated in this book will keep scholars busy for years to come.

The explanation of how these hypotheses are tested, however, is often opaque. Morrow tries to make the book accessible by providing two versions of several chapters: one without the technical details and another with them. However, even empirically sophisticated readers will have difficulty following exactly how Morrow tests each of his hypotheses. For example, Morrow concludes his theory chapter by listing seven clear hypotheses that are derived using game theory (pp. 86-88). He then starts the first version of his empirical chapter — the version meant to be accessible to all readers — by listing six questions that he plans to address (p. 111). These six questions are similar, but not identical, to his seven hypotheses. The second version of his empirical chapter — the version meant for quantitatively sophisticated readers — lists five questions that the chapter will address (p. 166). This structure and organization can make it difficult to evaluate the methods used to test, and the evidence in support of, Morrow’s hypotheses.

Moreover, although Morrow’s predictions do not turn on “regime type” (whether a country is a democracy or not), he begins his analysis of the data with countries broken out by regime type. Although may other scholars have argued that regime type is an important predictor of compliance with international law, and Morrow provides an explanation of his decision to present the data this way (pp. 113-116), his empirical analysis is complicated by beginning with a conditional theory that is not directly derived from the theory chapter.

Another issue is how generalizable Morrow’s theory will be to future conflicts. Morrow’s theory is developed based on the behavior of states during the World Wars. Since that time, many of the relevant treaties have achieved near universal ratification. For example, the primary post-war treaty on the laws of war — the Geneva Conventions of 1949 — has been signed by 196 countries. If, as Morrow argues, treaty negotiation and ratification reveal information about countries’ preferences, then this information will become stale as the ratification period recedes in time.

Additionally, at the same time that ratification of these treaties has become widespread, interstate wars have become less common. This is not simply because there are no longer armed conflicts. Instead, wars within countries have been more common than wars between countries. Given his exclusive focus on interstate wars, Morrow’s book may explain more in the last century than it will in the next.

If Morrow’s theory is correct, however, it does have important implications for how we should think about the development of the laws of war. Morrow’s theory is not that the laws of war are important because they change states’ preferences, but instead that they are important because the negotiation and ratification process reveals those preferences. If this is true, maybe advocates should put more energy into persuading countries to negotiate new treaties (say, for example, on the use of drones or cyberattacks). These negotiations can then serve to reveal what restrictions states are willing to subject themselves to during conflicts. Morrow’s theory also suggests that when these new treaties are developed, it would be a mistake to strong-arm countries into ratification because it would degrade the information value of ratification. This approach stands in contrast to the way that many advocates focus on promoting the ratification of treaties that already exist and on compelling countries to live up to their treaty obligation.

Morrow’s work also matters for how we should understand other areas of international law. If, as he argues, compliance with international treaties is largely a matter of states cooperating through the threat of reciprocal sanctions, treaties that create public goods (like a climate treaty) and advance global values (like human rights treaties) are hard to explain.

Order Within Anarchy is a theoretically rich analysis of the laws of war that will inspire more scholarship on this important topic. A considerable amount of diplomatic effort was spent during the twentieth century drafting treaties to help reduce the horrors of war, and Morrow’s book is an impressive attempt to figure out whether those efforts have worked.

ADAM CHILTON is an Assistant Professor of Law at the University of Chicago Law School. His research focuses on using empirical methods to study international law.

[1] It total, Morrow generates seven hypotheses (pp. 86-88).

[2] Interstate wars are armed conflicts between two or more countries. Intrastate wars, on the other hand, are armed conflicts between two or more parties within a single country.